SNA News and Notes

Issue 8, November 1998

• SNA68 and SNA93 Compared: Public Enterprises
• What is FISIM?
• Canada publishes report on historical revision of Canadian SNA
• 1993 SNA Implementation in the Philippines
• Treatment of interest under high inflation
• Status of SNA News and Notes Articles
• Manual on Monetary and Financial Statistics
• Treatment of Asset Transfer Costs
• Editorial Note
  
  

SNA68 and SNA93 compared: Public enterprises

By Paul McCarthy, OECD (For Information)

Question: To what extent has the definition of the public corporate sector changed in SNA93 compared with the old SNA?

Answer: In practice, general statistical compilationsthe scope of public enterprises has not really changed at all as a result of the introduction of SNA93. However, the terminology used has changed and it is possible that this could cause some confusion. Under both frameworks, the critical issue in determining if a business is a public enterprise is whether or not an arm of government controls the business. In answering the question of who controls a business, it is useful to look at some of the detailed descriptions in each of the old and new SNAs.

SNA68

Paragraphs 5.10, 5.11 and 5.12 define the characteristics of what the 1968 System refers to as "industries of government" or "public enterprises", and what the 1993 System refers to as "public corporations". The key element here is the first sentence in paragraph 5.10 which reads "Industries should encompass government departments, establishments and similar units mainly engaged in selling the kinds of goods and services which are often produced by business establishments though as a matter of policy, the prices (charges) set for these goods and services may not approximate the full costs of production." Paragraph 5.55 is the key to determining whether or not a unit is public or private. The key points are shown in bold type (my emphasis): "The last classification is based on whether the ownership and/or control of an enterprise rests in the public authorities or private parties. The public authorities, or private parties, are considered to be the owners of a given enterprise if they own all, or a majority, of the shares of, other forms of capital participation in, or equity of the unit. The criteria required in order to determine who controls an enterprise are more complex… The important consideration in determining whether the public authorities are in control is: do they exercise an effective influence in all the main aspects of management; not merely such influence as is derived from the use of regulatory powers of a general kind."

The key issue that needs to be noted in this paragraph is in the last sentence, which refers to exercising an "effective influence in all the main aspects of management". It is a criterion which could result in an enterprise being classified as public even if the government did not have more than 50% of the capital in the enterprise (this would cover situations such as the so-called "golden share" where governments have sold off more than 50% of the shares but retain a veto right on decisions in respect of the management of an enterprise).

SNA93

The terminology in SNA93 is different in a number of ways from that in SNA68 but with no significant changes being made to the underlying concept. In particular, the control criterion in SNA93 is set out using different wording than in SNA68. While being much the same in practice, the SNA93 criterion also gives some specific examples of situations in which government can control an enterprise even if it has less than 50 per cent of the shares. Paragraph 4.27 states, in part, "...in general an individual institutional unit or group of units owning more than half the voting shares of a corporation can exercise complete control by appointing directors of its own choice".

Paragraph 4.30 further broadens this criterion. It states, in part, "...control is defined as the ability to determine general corporate policy by appointing appropriate directors, if necessary. Owning more than half the shares of corporation is evidently a sufficient, but not a necessary, condition for control. Nevertheless, because it may be difficult to identify those corporations in which control is exercised by a minority of shareholders, it is recommended that, in practice, corporations subject to public or foreign control should normally be confined to those in which governments or non-residents own a majority of the shares. This recommendation is intended only as a practical guideline, however, to which exceptions can be admitted if there is other evidence of control. For example, a corporation which government is able to control as a result of special legislation should be treated as a public corporation even if the government does not own a majority of the shares." The highlighted part of this paragraph indicates that a business can be classified as a public corporation even if the government owns less than 50% of the shares, as is also the case in the 1968 System. In addition, paragraph 4.24(f)(i) is relevant because it also indicates that governments may exercise control without a majority of the shares ("The voting rights of shareholders may not be equal. Some types of shares may carry no voting rights, while others may carry exceptional rights, such as the right to make specific appointments to the board of directors or the right to veto other appointments made on a majority vote. Such exceptional rights may be held by the Government when it is a shareholder in a corporation").

This paragraph is a formal description of the "golden share" referred to above.

Conclusion

Public "corporations" or "enterprises" are very similarly defined in both SNA93 and SNA68 because both frameworks consider that any business over which a government has effective control should be classified as a public enterprise. Both SNA68 and SNA93 indicate that holding more than 50% of the shares is a sufficient condition (SNA68 by stating the condition as a majority of the shares and SNA93 by stating it as more than half the voting shares). In addition, both indicate that enterprises in which government holds less than half the shares may still be classified as public corporations (SNA68 in paragraph 5.55 and SNA93 in paragraphs 4.30 and 4.24(f)(i)) provided that government nevertheless controls the business through an influence on all main aspects of management.

What is FISIM?

By Robin Lynch, World bank ( For Information)

 This short article sets out in simple terms what FISIM is, and the effect on the level of GDP of replacing the financial services adjustment of the 1968 SNA with FISIM (Financial Intermediation Services Indirectly Measured) allocated to users according to the 1993 SNA. Banks make a large part of their money by lending at a high rate of interest, and borrowing at a low rate. The difference between the two, results in banks receiving net interest receipts. Because this income arises from the difference between interest rates, banks do not need to charge directly for all the services they provide in arranging borrowing and lending. For example, checking accounts are usually maintained free by the banks, and the associated costs are met by the difference between the low interest payments awarded on credit balances maintained in such accounts whilst the bank lends it at a higher rate to a borrower.

There is a notional reference rate of interest at which lending and borrowing can take place directly between a lender and a borrower, satisfactory to both parties. So consider if I want to borrow 100 pesos and I am prepared to pay back 105 pesos after one year.

If I could find the right person at the right time, it's quite possible that they would accept this as good business, and lend me 100 pesos in exchange for the 105 in a year's time. However, as we all know, finding a counterpart who would be interested at the same time in the same deal over the same time period isn't easy, so we go to a bank to help us. They bundle together various lending offers, and match them against the array of borrowers so that they effectively provide a market for borrowing and lending. If in order to make a living out of such transactions, it turned out the bank had to charge 5 pesos, then they could choose to directly charge 5 pesos for the introduction and facilities to make the transaction. If the bank does it through the interest rates, then in the example above, I could be asked to pay back 108 pesos at the end of the year, but the lender might receive only 103 pesos. The net interest receipts of the bank are equal to 8 less 3 = 5 pesos. If we accept that the reference rate of interest for this kind of small borrowing and lending is 5%, then we can see that I have paid 3 pesos, and the lender 2 pesos.

In the national accounts, interest payments are not considered to be payments for a service rendered, but a form of property income, and so are recorded in the generation of primary income account, and not the production account. But because of the peculiar way that banks make their money, by loading the interest rates, this makes it seem that the banks are given all the money for their services in the income accounts, rather than "earning it" for services provided in the production account. So the operating surplus of banks would show as a negative item, given that most of the income earned from borrowing and lending would appear later in the current income and expenditure account. As this would give a false impression of the size of the operating surplus of banks compared to other industries, the system of national accounts adopted the following solution. Let the output of the banks include net interest receipts as if customers paid for the service (this service is what we label "FISIM.") But no such payments are shown in the production accounts of industry or the final consumption expenditure of households.

To be consistent with showing the net interest receipts contributing to the bank output, we should show the corresponding intermediate consumption of banking services in the production accounts of the industries using the banking services, and the final expenditure in the appropriate component of final demand. But we don't know with confidence which industries and final consumers are making the payments. So the 1968 SNA invented a nominal industry with a very strange production account - it had no output, and one intermediate input - all the FISIM payments by all the bank customers - businesses, government, the rest of the world and households. As a consequence, the operating surplus of this nominal industry was equal to the negative figure of the net interest earned by banks. This device had the effect of reducing gross domestic product by the amount by which it had been increased by including net interest receipts in the value added of the banking industry in the first place. But there is still a snag. Part of the financial service is paid by final consumers, and not consumed in connection with a production process, and the above treatment ignores this. So in reducing total value added by all the net interest receipts, we have gone too far. We should really have only taken off the net interest receipts paid by enterprises. But given the difficulty in identifying how much of the net interest receipts were paid by final consumers as opposed to industry, it was felt that a reasonable compromise had been struck between ignoring it completely, and making a very shaky allocation of the payment of net interest receipts among industry and final demand components. So it was left as an anomaly that GDP measured according to the 1968 SNA was too low by the amount of financial service payments imputed to non-business consumers through the loading of interest payments by banks.

The situation is illustrated in the simple use matrix shown, where the columns for industry represent the production accounts. Consider an economy with only two industries - banks and the rest, and only private consumers in final demand. Let's suppose that the banks load the interest rates so that business pays 20m pesos indirectly for the financial service provided, and consumers pay 30m pesos. So the operating surplus of the firms includes 20m pesos which it pays to banks through the interest flows in the current account. Similarly, consumers are shown as not spending anything on financial services as consumers' expenditure, as they "spend" it through the interest flows contained in the current account for the households. In the simple example, banks are assumed to make all their money through net interest receipts.

Table 1. Interest payments are treated solely as property income 

If we could identify all the service payments with confidence, and treat them as such, not interest payments, then the picture would be as follows.

Table 3. 1993 SNA treatment of interest payments 

This illustrates that the 1968 treatment results in an understatement of total value added by the amount paid by final consumers (Table 2). The 1993 SNA suggests that we follow the theoretically correct treatment, that is, according to the third method (Table 3). This implies that in comparison with the 1968 treatment, in the primary income account the figure of interest paid by borrowers goes down (because the FISIM is taken out). Conversely, the figure for interest received by lenders goes up (because the interest receipts now exclude payments of FISIM to the banks).

In order to estimate the correct level of GDP, it is necessary to make the allocation of the service payments across industries and consumers with confidence. For each transaction where there is a loading of interest payments to cover the costs of the financial services provider, we would like to know who the borrower is, who the lender is, what the lending instrument is, and what is the appropriate reference rate for the transaction. In practical terms, we cannot know this detail and must settle for more aggregate measures. The technique of estimating the appropriate reference rates and for what level of aggregation of users and financial instruments, has been the subject of debate and ongoing experiment, and the outcome is still uncertain in practical terms. European Union members will not take a decision on whether to allocate FISIM until the year 2002, when more is known about the effect of the various allocation options in practice.

Canada publishes report on historical revision of Canadian SNA

By Kishori Lal , Statistics Canada (For Information)

As already announced in a feature article in the SNA News and Notes No. 5 (January 1997), Canada has now completed a historical revision of its national accounts. In this context, four documents have been published recently.

The first one "The 1997 Historical Revision of the Canadian System of National Accounts, Record of Changes in Classification of Sectors and Transactions, Concepts and Methodology", elaborates on the 102 changes made in the accounts. This is an update to an earlier report issued in August 1996. The second report has documented the changes made to actual statistics for one of the most important SNA series, the current price Gross Domestic Product (GDP) for the period 1961-93, "A Statistical Representation from the Old to the New". It includes 13 tables detailing various GDP components.

 The third report is entitled "Industry, Final Demand and Commodity Classification Systems". This report details how dealing with various SNA classification systems is crucial for intertemporal comparisons.

The fourth report sets forth and explains the "Remaining Differences between the 1997 Canadian System of National Accounts and the 1993 International System of National Accounts". It is planned to produce several other documents during the next two years, which will set forth and explain remaining differences between international Balance of Payments standards and current Canadian practices; propose changes in the 1993 SNA to be reflected in further versions and, document subject-specific sources and method.

Copies of the above mentioned documents are available upon request from my office:
System of National Accounts, Statistics Canada,
Ottawa, Ontario, Canada K1A 0T6,
Fax 1.613.951-9031,
e-mail: lalkish@statcan.ca

 1993 SNA Implementation in the Philippines

By Estrella V. Domingo, (For Information)

National Statistical Coordination Board

As a result of the cooperative efforts with the UNSD and the ESCAP the National Statistical Coordination Board (NSCB) of the Philippines is currently piloting a project on the Institutional Strengthening of the Philippine System of National Accounts (PSNA). The Philippines' practice could serve as a reference for the Asian and Pacific Region on the implementation of the 1993 SNA.

The general objective of the Project is to implement and institutionalize the PSNA according to the recommendations of 1993 SNA. Meanwhile the following specific objectives were set: integrate the various accounts into a consolidated system of economic accounts; develop estimation procedures and compilation systems that are compatible with the available supporting data and computerize the process; upgrade the capability of the compilers of the accounts and enhance the participation of data producers in the SNA; revise the estimates for the years 1994-1997; and document the experience. The first step in the project is to compile supply and use table and accounts up to the balance sheet for the years 1998 to 1999, using the currently available data, meanwhile the rest of the accounts will be operationalized after l999. For the sake of comparability, until 2001 the statistical agency is to continue compiling the national accounts according to 1968 SNA as well. For the time period of 1994 to be 1997, the above mentioned accounts are to be compiled in 1998 and 1999, applying the 1993 SNA concept. In 2000 and 2001 a ten-year series of the PSNA will be compiled according to the 1993 SNA. The full implementation of the 1993 SNA is expected for 2002. The following major activities will be undertaken:

1. Development and Operationalization of the 1993 SNA. This step is to include activities like establishment of mechanisms to link up with data source agencies, creation of inter-agency task forces, strengthening the system of designated statistics, collaboration with the tax agency, etc. At the same time special studies will be conducted (e.g. on FISIM and Insurance Service Charges, Identification/Measurement of NPISH; Linking Establishment and Enterprise Data; Chain Indices; Taxes; Transfers in Kind; Developing Household Account).

2. Review of Existing Databases and Surveys. The review will set the direction for data improvement program to support the 1993 SNA.

3. Development of a Computerized Compilation System. This is intended to systematize the compilation process for the PSNA. Intermediate worksheets will be designed to facilitate transfer of data from data sources to NA compilers.

4. Country Courses. Two country courses have been conducted. While the first training was on concepts and theories on the 1993 SNA, the second training dealt with the compilation process.

5. Preparation of publications on methodology. To assist the Philippines and other countries in the future, the Philippine experience will be documented through training manuals and sources and methods manuals.

6. Inter-Regional Workshop.

7. Preparation of a Blue-Print for Institutionalization.

Treatment of interest under high inflation

By Cristina Hannig, UNSD (For Information)

 It has been brought to the attention of the ISWGNA that there appears to be an inconsistency in the 1993 SNA regarding the treatment of interest under high inflation.

 The ISWGNA has discussed the controversy and has agreed to initiate a process of consultation with a panel of experts in this field in order to elucidate the exact nature of the problem and make recommendations as appropriate.

Status of Sna News and Notes articles

By Stefan Schweinfest, UNSD (For Information)

 At a recent meeting the Inter-secretariat Working Group on National Accounts (ISWGNA) discussed the status of articles published in the SNA News and Notes. On the one hand it was recognized that given its large readership the newsletter was an excellent vehicle to disseminate official positions of the ISWGNA, for instance with regard to clarifications or interpretations of the SNA text. On the other hand it was also felt that the newsletter should remain an open forum for debate - sometimes controversial - and exchange of information. In order to avoid confusion, SNA News and Notes will clearly mark in the future, which articles reflect an official position of the ISWGNA. Articles which are meant to invite for an open debate will be marked 'Discussion' and articles that inform on a specific country experience or on other issues relevant for SNA implementation will be labeled 'Information'. The table below takes a look back at contributions that were published in the previous 7 issues of SNA News and Notes and offers a classification of the article.

 Table 4. Categorization of articles published in the previous issues of SNA News and Notes

 Manuals on Monetary and Financial Statistics

By Adriaan Bloem , IMF (For Information)

 The Manual on Monetary and Financial Statistics (MMFS), which will provide guidance on the compilation of monetary statistics and a broad range of financial statistics, is now being prepared for publication. In November 1996, the Expert Group on Monetary and Financial Statistics, which comprised experts designated by member countries and representatives of regional central banks and international statistical organizations, was convened to review the MMFS. The Group strongly endorsed the methodology and presentation of the MMFS and gave its approval for publication. The MMFS is now undergoing a final review; its publication in English is expected around year end. Publication in other languages is expected beginning in 1999.

Work commenced in early 1995 on the revision of A Manual Government Finance Statistics, 1986 (GFS), which sets standards for compiling statistics used in fiscal analysis and in reporting of government finance statistics to the Fund. Views of member countries were solicited on changes that could be made to the GFS Manual. An English-language version of an annotated outline of the proposed revised Manual was sent in September 1996 to member countries and other reviewers. Dispatch of non-English versions of the outline was completed in February 1997. Major changes to the GFS methodology, including a replacement over time of the predominantly cash basis of recording with a predominantly accrual basis, are being proposed. The new system will include full balance sheets that reconcile, in an integrated and transparent manner, all transactions, revaluation, and other economic events with balance sheets. The new approach will permit the development of GFS reporting on an accrual basis as well as on a cash basis. Drafting of the revised GFS Manual has proceeded on the basis of comments received on the annotated outline. Several significant issues of methodology and terminology that remained unresolved are currently being addressed.

Treatment of asset transfer costs

By Paul McCarthy, OECD (Discussion)

 SNA93 recommends that the assets of transferring the ownership of existing non-financial assets should be treated as gross fixed capital formation (GFCF) in the national accounts (paragraph 10.57 of SNA93).

Broadly, ownership transfer costs include all professional charges and commissions incurred by either the buyer or seller, plus all taxes payable by either party (paragraphs 10.55 and 10.56 of SNA93). At a joint OECD/ESCAP national accounts meeting in May 1998, the Singapore Department of Statistics (SDOS) presented a paper identifying some problems, which arise when the SNA93 recommendations are followed. In summary, they are: It is difficult to visualize the "asset" being created by the transfer process – e.g. is an asset, which changes hands several times more valuable or productive than an identical asset, which has remained under the same ownership for its entire life?

The recommended treatment of capitalizing these transfer costs is different from that recommended for transfer costs associated with transfers of financial assets (paragraph 11.44 of SNA93). To the extent that real estate transactions are involved both GFCF and saving may be artificially inflated during property booms, as a result of speculative activities.

The solution proposed by SDOS is to treat transfer costs as current expenses rather than as capital expenditure. The ISWGNA considers the SDOS proposals need to be investigated in more detail. Therefore, it is initiating consultations on this issue. Any views should be sent to the UN Statistics Division (see Editorial Note).

Editorial Note

SNA News and Notes is a bi-annual information service of the ISWGNA prepared by United Nations Statistics Division (UNSD). It does not necessarily express the official position of any of the members of the ISWGNA (European Union, IMF, OECD, United Nations and World Bank)

SNA News and Notes is published in four languages (English, French, Russian and Spanish) and can be accessed on the internet: http://www.un.org/Depts/unsd

Correspondence including requests for free subscriptions should be addressed to:
UNSD, Room DC2-1720,
New York, NY 10017;
tel.:+1-212-963-4854,
fax:1-212-963-1374,
e-mail: sna@un.org.